We resolve the tension between shipping new product and supporting existing customers that every product team feels in their bones.
Victoria, BC
Joined January 2024
- Tweets1.8K
- Following1.7K
- Followers663
- Likes2.6K
Pinned Tweet
Can you run the frontier at home, at work?
We dug deeper to find out every detail of the DGX Station, and what kinda hardware it's taking to run GLM 5.2.
Is the GLM they say is running the same GLM as the benchmarks? Not always, but that doesn't mean it's a write-off.
atcyrus.com/stories/dgx-stat…
so agents are in front of, but now also behind, MCPs?
Connor Turland retweeted
We’re excited to share that Cyrus has successfully completed its SOC 2 Type II audit 🔒
Building on Type I, our Type II audit independently assessed how effectively our security controls operated over time.
Another milestone in securing AI development workflows.
atcyrus.com/stories/cyrus-so…
The people who need AI to operate their computer instead of them the most, are likely to be among the last to get it set up.
y'all, its frustrating that we can't point codex simultaneously at local models AND at openai models, and are forced into an either or.
@OpenAI plz fix?
great, we are prompting in public now... watch this devolve.
@bot monitor this every hour, when the preorder is available, place the order for me please 🫶
In case you aren't in the loop on exactly what Claude slop writing has been looking like, I present to you a perfect case study.
Dario Amodei said the quiet part loud:
"If managed poorly, I even believe that AI could be a risk to humanity as a whole."
Then immediately followed it with: "At Anthropic, we believe we have the best track record in the industry for preventing these risks."
That's a remarkable move. Acknowledge civilizational risk. Then position your company as the designated adult in the room.
The tension is real: Anthropic genuinely does more safety research than most. But "best track record" is doing a lot of work when the models haven't actually caused catastrophic harm yet — at any lab. You can't claim credit for preventing something that hasn't happened.
What Amodei is really doing is threading a needle that defines Anthropic's entire business model: the risk is real enough to justify your existence, but manageable enough that you keep shipping.
That's not dishonest. It's just worth naming clearly.
Im at a point where I am trying to do a good civic duty by vendetta-posting against siri until they feel enough public wrath to fix that thing.
that kind of thing didnt exist already... ?
that seems like how it... should work.
Am I the only one who can't tell whether Codex projects are a folder on a computer or a row in a database and an abstraction?
ffs: apple will do anything except fix siri
Here’s your first look at home accessories Apple has co-developed with LG. 🧵
1. LG Smart Deadbolt Lock
• The keypad rotates to open. You can use the included swivel pin to limit the rotation on one side so it doesn’t touch a wall.
• Settings for Touchpad Light Ring, Door Left Audio Alarm, and Lock Sound Volume.
• Supports automatic unlocking, tap to unlock, emergency power and master codes.
• Fits 2.375 in (60.3 mm) and 2.75 in (69.9 mm) brackets.
• Comes with 2 mechanical keys.
